+ MODEL

Available online at www.sciencedirect.com

Borsa _Istanbul Review

_
Borsa Istanbul Review xxx (xxxx) xxx
http://www.elsevier.com/journals/borsa-istanbul-review/2214-8450

Full Length Article

Internal and external corporate social responsibility activities and firm value:
Evidence from the shared growth in the supply chain
Woo Jae Lee a, Seung Uk Choi b,*
a
Department of Accounting, College of Commerce, Jeonbuk National University, 567 Baekje-daero, Deokjin-gu, Jeonju, 54896, Jeollabuk-do, Republic of Korea
b
Department of Accounting and Taxation, School of Management, Kyung Hee University, 26 Kyunghee-daero, Dongdaemun-gu, Seoul, 02447, Republic of Korea
Received 29 September 2020; revised 1 February 2021; accepted 1 February 2021
Available online ▪ ▪ ▪

Abstract

Despite the long history of corporate social responsibility (CSR) research, few studies have focused on CSR activities related to business
partners in supply chains. In this regard, we investigate whether internal (or backward) CSR enhances firm value. Using an index from the Korea
Commission for Corporate Partnership, which indicates whether a firm shares its profits with business partner companies, we find that firm value
increases as this backward CSR increases. Further, after controlling for internal CSR, firms engaged only in external CSR have lower firm value
than non-CSR firms. Interestingly, we find that firms involved in both internal and external CSR have higher value. Overall, by showing that
internal CSR is a core activity that enhances firm value, our study provides policy implications for the regulatory bodies of different countries.
_
Copyright © 2021, Borsa Istanbul Anonim Şirketi. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

JEL classification: M14; M48

Keywords: Internal CSR; External CSR; Firm value; Supply chain; Shared growth

1. Introduction Previous CSR studies finding an association with firm value

This study investigates whether corporate social re-
sponsibility (CSR) practices related to both suppliers (internal
CSR) and the community (external CSR) positively affect firm
value. As prior research focuses primarily on the value of
external CSR activities, we expand the literature by examining
the effect of internal CSR and the convergence of internal and
external CSR on firm value.
CSR research has a very long history. Prior research cate-
gorizes CSR into economic, legal, ethical, and discretionary
responsibility (Carroll, 1979). CSR is the extent to which firms
assume the four types of responsibilities toward the stake-
holders in their society (Maignan, Ferrell, & Hult, 1999).
* Corresponding author.
E-mail addresses: wjlee@jbnu.ac.kr (W.J. Lee), suchoi@khu.ac.kr (S.U.
Choi).
_
Peer review under responsibility of Borsa Istanbul Anonim Şirketi.
indicate that strategic CSR practices provide the firm with
resources that induce positive behaviors among stakeholders,
leading to higher market value. Prior research also argues that
firms might gain competitive advantage by conducting social
activities for external stakeholders (Porter, 1996), which helps
increase their value. Therefore, prior studies suggest that
external CSR enhances firm value (Harjoto & Jo, 2015; Cahan,
De Villiers, Jeter, Naiker, & Van Staden, 2016). In contrast,
other studies argue that external CSR is a burden that harms,
or is not related to, firm value. The critiques are based on the
view that firms use CSR only to create a positive image
without substantially improving the firm (Marquis & Qian,
2014). From a similar perspective, recent empirical studies
show a drop in the value of firms investing in external CSR
(Lenz, Wetzel, & Hammerschmidt, 2017; Manchiraju &
Rajgopal, 2017).
Regarding internal CSR, there is a lack of empirical evi-
dence on how CSR activities targeting suppliers affect a firm's

https://doi.org/10.1016/j.bir.2021.02.007
_
2214-8450/Copyright © 2021, Borsa Istanbul Anonim Şirketi. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: W.J. Lee, S.U. Choi, Internal and external corporate social responsibility activities and firm value: Evidence from the shared growth in
_
the supply chain, Borsa Istanbul Review, https://doi.org/10.1016/j.bir.2021.02.007
 + MODEL
W.J. Lee, S.U. Choi
value. Strong relationships with suppliers can build mutually
profitable business connections (Kwon & Suh, 2005). Despite
the long history of CSR research, implementing CSR in supply
chains is relatively uncommon. Prior research on this issue
uses a small sample of observations focusing on a case study
of a specific firm or industry (e.g., Lindgreen, Swaen, Maon,
Andersen, & Skjoett-Larsen, 2009; Maloni & Brown, 2006).
Therefore, we expand the literature by investigating the rela-
tionship between CSR in the supply chain and firm value.
Sharing firms' benefits with all suppliers in the supply chain
has been one of the most important issues in Korean society in
the last two decades. The Korean government expects mutual
growth among business partners to be the key to overcome the
country's stagnant growth and to move from a developing
country to a developed one. Consequently, the Korea Com-
mission for Corporate Partnership (KCCP) was founded in
2011. The KCCP evaluates firms with strong influence on
society once a year. It computes and publishes its Win-Win
Growth Index (WWGI) to promote partnerships between
large and small companies in the same supply chain. The
WWGI consists of the following subcategories: the level of
fairness of the contract and trade made between business
partners as an effort to build a system monitoring law viola-
tion; technical and human resource support for work condi-
tions; and sharing visions and information. By regularly
evaluating the shared growth level for each firm, the com-
mission encourages firms to share the benefits with all sup-
pliers by mutual growth. The WWGI has been published since
2011 based on the Act on the Promotion of Collaborative
Cooperation between Large Enterprises and Small-Medium
Enterprises. The evaluation and publication of the WWGI
are under the supervision of the National Fair Trade Com-
mission. We use a dichotomous variable indicating whether a
firm is included in this index as a proxy for its internal CSR
activities. We also use the internal CSR scores to identify the
effectiveness of these activities.
Next, external CSR is proxied by the index from the Korea
Economic Justice Institute (KEJI), which provides a detailed
CSR scores of listed firms since 2000. The “Best Corporate
Citizen Index” awards disclosed by KEJI include the scores of
the top 200 firms with highest CSR scores. The index includes
the following sub-indexes: soundness, fairness, social contri-
bution, consumer protection, environmental management, and
employee satisfaction. Many prior studies on CSR using South
Korean firms adopt this index (e.g., Chang, Oh, Park, & Jang,
2017; Park & Ha, 2020).
The sample consists of firms listed in the Korean stock
market from 2013 to 2018. Using the KCCP's internal CSR
index and the KEJI's external CSR index, we first analyze the
effects of internal and external CSR on firm value separately.
The results of this within-sample analysis (examining each
CSR index separately) show that increased internal CSR
scores increase firm value as measured by Tobin's Q. We also
find a positive relationship between external CSR and firm
2
_
Borsa Istanbul Review xxx (xxxx) xxx
value. The results are robust when using the current or one-
year-ahead values of Tobin's Q or when using the industry
mean-adjusted Tobin's Q. Furthermore, the results from the
full sample analysis indicate that firms involved in both types
of CSR or that invest in only internal CSR have higher value
than do non-CSR firms. In contrast, firms engaged in only
external CSR have lower value than non-CSR firms. Thus,
social activity for internal suppliers is a key factor determining
the effectiveness of CSR activities for firm value. Our results
are robust when two-stage regressions or instrumental ap-
proaches are used to mitigate endogeneity concerns in CSR
and firm value. Lastly, we find that the positive effect of in-
ternal CSR on firm value is robust after controlling for the
promotional donation to suppliers.
Our study offers several contributions to the literature. First,
there is limited empirical evidence on backward or internal
CSR. By applying a fine-grained analysis using the KCCP's
CSR data related to supply chain activities, this study contrib-
utes to the literature by highlighting supply chain implications.
Our findings suggest that the positive effect of CSR activities
related to suppliers on their innovation, which influences the
firm's production differentiation, leads to an increase in firm
value. Second, a major contribution of our study is the finding
that the positive relation between CSR and firm value derives
mainly from internal CSR activities. These results are valuable
in explaining the specific path through which social activities
increase firm value. By studying the internal supply chain, we
provide clearer, more direct results indicating that social con-
tributions are connected to firm value. Third, prior research on
external CSR and firm value presents mixed evidence. Our
findings suggest that it is important to consider internal CSR
activities when testing the effect of CSR on firm outcomes.
Our study has important policy implications. Emerging
countries with substantial development, such as Korea, have
reached a state of constant low growth. To overcome this
stagnation, the Korean government is making efforts to induce
communal growth via profit sharing among the business
partners in a supply chain. This is a new paradigm of social
contribution and an activity directly affecting the growth and
profits of firms. By showing that a firm's internal CSR activ-
ities increase its value, our study confirms that these social
efforts are successful. Moreover, our results indicate that profit
sharing can help an economy reach a socially desirable state,
suggesting that regulators need to supplement these achieve-
ments through proper regulation. Thus, the results also have
policy implications for the regulatory bodies of different
countries. Lastly, we call for future research to investigate the
ways to continue internal CSR activities and their effective-
ness on other firm outcomes.
The remainder of the paper is organized as follows. Section
2 presents a review of the literature on CSR activities and
develops the hypotheses. Section 3 describes the research
design and sample. Section 4 reports the empirical results, and
Section 5 concludes the study.
 + MODEL
W.J. Lee, S.U. Choi
2. Literature review and hypotheses development
2.1. Effects of customer-facing CSR activities (external
CSR) on firm value
Despite the long history of research on the effect of CSR on
firm outcomes, it remains unclear empirically whether
investing in CSR leads to a higher firm value. For instance, a
large body of research based on stakeholder theory finds that
the firm's characteristics affect the consequences of CSR ac-
tivities. For instance, a stream of research insists that a good
ownership structure, efficient corporate governance mecha-
nisms, and strong legal enforcement affect the effectiveness of
CSR on firm value (Cahan et al., 2016; Nekhili, Nagati,
Chtioui, & Rebolledo, 2017; Buchanan, Cao, & Chen,
2018). Cahan et al. (2016) document evidence that firms in
countries with strong legal enforcement to promote CSR dis-
closures show higher value than those in countries with rela-
tively weak enforcement. Buchanan et al. (2018) demonstrate
that the effectiveness of CSR on firm value varies with the
level of influential institutional ownership. In particular, the
authors find that the positive effect of institutional ownership
is more pronounced during a financial crisis. Nekhili et al.
(2017) suggest that as a part of the efforts to build trust with
stakeholders, family firms are more prone to engage in CSR
activities than non-family ones. Their study finds that this
moderating role of family owners leads to higher market value
among firms investing in CSR.
In contrast, some critiques based on stakeholder theory
suggest that firms use CSR only to create a positive image
without making practical improvements within the firm
(Marquis & Qian, 2014) or that firms use CSR with the intent
to reduce or divert attention away from the firm's misconduct
(Du, 2015). These are the perspectives shared by critiques
based on stewardship or agency theory. For instance,
Bhardwaj, Chatterjee, Demir, and Turut (2018) find that in-
vestment in CSR that does not directly improve the produc-
tion efficiency or the quality of a product, thereby cannot
enhance a firm's operating performance. Moreover, Lenz
et al. (2017) show that involvement in corporate social irre-
sponsibility may decrease a firm's value. Manchiraju and
Rajgopal (2017) report evidence that Indian firms that must
spend a certain portion of their profits on CSR activities by

law experience a drop in their value. Gallego-Alvarez, Prado-
Lorenzo, and García-Sanchez (2011) explore the negative
bidirectional relationship between CSR practices and the
innovation within the firm, suggesting that not all CSR ac-
tivities improve firm value. Similarly, Bhandari and
Javakhadze (2017) show that investing in CSR reduces a
firm's resources, therefore causing investment inefficiency.
Overall, despite the theoretical support that CSR may posi-
tively affect firm value, the empirical evidence from previous
studies vary depending on the theory, firm, country, and
method of analysis.
3
_
Borsa Istanbul Review xxx (xxxx) xxx
2.2. CSR practices to backward suppliers (internal CSR)
and firm value
This study focuses primarily on the backward (or internal)
CSR activities, which are the CSR activities targeting the
business partners within the supply chain. By engaging in
CSR, firms will also engage in activities that benefit both
backward and forward stakeholders, who contribute to the
firm's productivity by supporting and endorsing them and
developing enduring relationships (Bosse & Coughlan, 2016).
The CSR activities do not have to be visible, but specifically
benefit the backward stakeholders, such as suppliers. For
instance, a financial capital company that invests heavily in
CSR may voluntarily provide reports to reduce or remove
agency costs and information asymmetry, and at more favor-
able terms, may use the resources from debtholders, institu-
tional investors, and other financial capital providers (Cheng,
Ioannou, & Serafeim, 2014). In the manufacturing industry,
raw material suppliers profiting from a firm's CSR activities
will also attempt to reciprocate by creating more productive
and cooperative relationships with the firm. Similarly, adopt-
ing and conducting CSR activities targeting suppliers can help
firms advance their existing innovation capabilities and create
new ones.
The presence of substitutable or complementary innovation
assets and the degree to which a firm absorbs the innovation
capabilities will promote the implementation of CSR activities
related to its assets. The relationships between absorptive
capability, innovation, and firm value have been well explored
(Tsai, 2001). However, these relationships were rarely
explored in the context of firms’ partnerships with suppliers.
Unlike the long history of external CSR research, the inves-
tigation of CSR related to supply chain partners began only
recently, in the early 2000s, and increased consistently in the
last two decades (Feng, Zhu, & Lai, 2017). Most studies were
published in the past five years. Therefore, research in this
field is a recent trend in the CSR literature and is highly likely
to be developed in the future. We summarize recent CSR
studies on supply chains as follows.
First, an increasing number of studies report how firms
conduct CSR in a specific industry for their supply chains.
For instance, some studies report the consequences of CSR
practices for firm performance in the food industry (Maloni &
Brown, 2006) or in a specific company with a global supply
chain (Lindgreen, Swaen, Maon, Andersen, & Skjoett-
Larsen, 2009). Second, studies also examine the effect of
CSR on supply chains for small and medium-sized enter-
prises (SMEs). The results generally indicate that SMEs are
less engaged in CSR for their business partners. For instance,
using survey data from Danish SMEs, Pedersen (2009) shows
that relatively large SMEs and those with the ability to
control and maintain CSR activities implement them in their
supply chain. Similarly, Lee, Kwak, and Park (2017)
demonstrate that SMEs tend to invest in external rather
 + MODEL
W.J. Lee, S.U. Choi
than internal CSR. Collectively, studies on SMEs find that
these firms focus more on CSR activities for external stake-
holders to protect their image and promote their brand
reputation among customers. Third, some studies investigate
why firms invest in CSR for their internal business partners.
Welford and Frost (2006) find that CSR activity reduces a
business partner's costs, leading to the efficient use of raw
materials and increase in an organization's productivity.
Using a quantitative method, Hsueh (2014) reveals that it
benefits both the manufacturer and the wholesale firm in the
supply chain if the manufacturer invests in CSR activities and
charges the costs to the wholesale firm's product price.
Overall, existing research in this field is relatively insufficient
and requires more empirical studies.
2.3. Hypotheses development
Profitable relationships are built mainly on the economic or
financial aspects by which the transactions between the firm
and suppliers are profitable to each other. Our main question is
whether CSR activities related to suppliers increase a firm's
value. The life cycle of a company's product starts from the
raw materials delivered by the suppliers and continues to the
manufacturing process before delivery to the consumer. The
supply chain refers to the series of connected entities that
encompass all suppliers and manufactures, and sometimes
even consumers. We propose several reasons to predict that
CSR activities for internal supply chains increase firm value
based on the theory from prior research.
First, the suppliers' innovation, driven by the firm's CSR,
can improve the products or services moving from the
suppliers to the firm, which could affect product differen-
tiation in a competitive market. CSR practices targeting
suppliers can occur in the form of an investment to achieve
product differentiation. Second, from the resource-based
view of the firm, CSR practices targeting suppliers could
create the greatest opportunity for success through
improved or new products supplied to the company. Under
this theory, the intangible resources and capabilities are
most important to firm success (Castelo & Lima, 2006).
Third, financial and technical support in human resources
and for working conditions can build trust among business
partners. For instance, prior research demonstrates that
information sharing between business partners builds trust
between them (Kwon & Suh, 2005). Overall, trust con-
structed from sharing knowledge and supportive resources
should have a significant impact on firm value. Lastly, we
propose that investment in the supply chain includes
business partners' efforts to prevent violations of the law.
Establishing a system that can post-monitor legal violations
among the business partners would positively affect the
relationship between suppliers and the firm. Additionally,
fair contracts and trades would impact the product inno-
vation capacity (Maloni & Brown, 2006).
All these voluntary efforts for the business partners within
the supply chain may lead to improved innovation capacity,
4
_
Borsa Istanbul Review xxx (xxxx) xxx
which will in turn increase the firm's value. Based on the above
discussion, we form our first hypothesis as follows.
H1. Internal (backward) CSR for the supply chain is posi-
tively related to firm value.
We also predict that firms engaged in forward CSR activ-
ities targeting external stakeholders, such as consumers, re-
sults in higher firm value. Beyond the CSR related to the
functional material, parts, or services provided by the suppliers
to the firm, CSR for external stakeholders will lead to an in-
crease in the firm's equity value because strategic CSR prac-
tices provide resources that induce positive behaviors among
stakeholders and thereby increase market value. In this regard,
prior research presents evidence that increased CSR enhances
firm value (Harjoto & Jo, 2015; Cahan et al., 2016). Therefore,
we expect that CSR activities that are visible to stakeholders
will increase the positive effects of CSR on firm value.
However, as aforementioned, there are opposing views on
the role of external CSR on firm value. The critiques suggest
that CSR delivers only a positive image to stakeholders
without making material improvement in the core value of the
firm (Marquis & Qian, 2014). In stewardship or agency theory,
CSR is just one means to cover managers’ opportunistic
behavior (Du, 2015). Similarly, recent studies show that firms
investing in external CSR experience a drop in their value

(Gallego-Alvarez et al., 2011; Lenz et al., 2017; Manchiraju &
Rajgopal, 2017).
To summarize, the previous findings are mixed and suggest
that not all external CSR activities create firm value. The
different evidences reported in prior studies vary across
countries, firms, and regulatory environments. We therefore
propose a second hypothesis to test whether external CSR is
positively related to firm value in the Korean stock market. We
propose the second hypothesis as follows.
H2. External (forward) CSR is positively related to firm value.
The first and second hypotheses lead to the discussion on
the synergy effect of conducting both internal and external
CSR on firm value. According to resource-based theory, both
the final consumers and suppliers are resources that create firm
success. However, these elements are different in terms of
creating product differentiation. The suppliers play functional
roles, producing goods or services by creating and providing
the material, parts, or services to the firms; the objective
quality of the companies' products or services can be deter-
mined by the quality of the products from the suppliers. In the
final stage of the market, it is possible that consumers will
subjectively evaluate the products or services based on their
existing knowledge of them. This subjective perception may
lead consumers to generate their own corporate product at-
tributes based on their existing knowledge about the firm's
external CSR activities. Product differentiation could be
perceived by both the objective quality and the subjective
perception. Therefore, we argue that involvement in both
 + MODEL
W.J. Lee, S.U. Choi
internal and external CSR will enhance firm value. Our third
hypothesis is as follows.
H3. Firms engaged in both types (external and internal) of
CSR have higher value than those engaged in only one type of
CSR or not engaged in CSR at all.
3. Research design
3.1. Proxy of internal CSR: the WWGI
Sharing the firm's profit with other business partners within
a supply chain has been an important economic issue in the
Korean society over the past two decades. This was one of the
main economic policies of the Korean government to counter
the country's lower growth rate. Consequently, the KCCP was
founded in 2011 with the aim of identifying influential com-
panies in terms of their accounting income and size and to
encourage them to share their profits with their business
partners. As one of the means to achieve this, strong com-
panies in terms of profit shares are selected and announced
every year with the aim of stimulating economic growth. The
results of the assessment are published in the WWGI. Based
on the WWGI score, the included firms are divided into five
grades (Best, Excellent, Good, Normal, and Insufficient).1 It is
important to note that even companies with poor grades are
expected to have better relationships with partners than those
not included in the index.
The procedure to evaluate the total score consists of two
parts. First, the National Fair Trade Commission (FTC) re-
views the performance of large companies based on their
submitted data and conducts on-site due diligence to check the
relationship with business partners to create the scores. The
other part of the score is evaluated by the KCCP itself. To
calculate fair and accurate scores, the KCCP visits SMEs in
person to collect the survey results. The survey contains sup-
pliers’ (who are in a weak position relative to the large firms)
responses about the co-prosperity of large companies, specif-
ically related to whether the larger firm uses a standard con-
tract, applies the ratio of cash payment, establishes a system to
prevent legal violations, provides financial, human resources,
employment, and technical supports, and uses fair trade
agreements among business partners. Based on the two scores
evaluated by the FTC and the KCCP, the latter releases the
WWGI, a result of the sum of the two scores. According to the
evaluation, about 150e200 companies are included in the
index every year.2
Among the top domestic companies regarding sales, com-
panies with social interests and those with influence on society
are selected in the evaluation process. Thus, inclusion in the
1
However, the KCCP does not disclose the specific scores. Only a summary
of data that classifies the companies into the five levels is released to the
public.
2
For instance, 189 companies were assessed in 2018. Among them, 31, 64,
68, 19, and 7 companies were classified as Best, Excellent, Good, Normal, and
Insufficient, respectively.
5
_
Borsa Istanbul Review xxx (xxxx) xxx
evaluation indicates that the company is expected to contribute
strongly to shared growth. To test H1, we generate a proxy of
internal CSR (CSRInternal) by assigning a value between 1 and
5 to firms classified as Insufficient to Best, respectively.3 A
higher number indicates a stronger contribution to internal
CSR activities. To test hypothesis 3, we construct a dichoto-
mous variable indicating whether a firm is included in this
index.
3.2. Proxy of external CSR: the KEJI index
Since 2000, the KEJI has provided a detailed CSR index of
listed firms. Every year, the institute discloses the “Best
Corporate Citizen Index” containing CSR scores. The evalu-
ation method has improved over time. The KEJI started pub-
lishing CSR scores in six categories after 2010. The evaluated
and disclosed firms are those ranked in the top 200 in terms of
their CSR scores. CSR studies using South Korean data typi-
cally adopt this index (e.g., Chang et al., 2017; Park & Ha,
2020). Total scores consist of six CSR sub-indexes. Listed
firms are evaluated based on soundness (25 points), fairness
(20 points), social contribution (15 points), consumer protec-
tion (15 points), environmental management (10 points), and
employee satisfaction (15 points) (see more detail at http://
ccej.or.kr). We use the total CSR score (maximum of 100
points) as a proxy of external CSR (CSRExternal). We also
construct a dichotomous variable indicating whether a firm is
included in this index to test hypothesis 3.
3.3. Research model
We set Eq. (1) to examine H1 and H2:
FirmValue ¼ a0 þ a1CSR þ a2SIZE þ a3LEV þ a4CURR þ
a5INVREC þ a6CF þ a7ROA þ a8LOSS þ
a9SalesGRW þ a10ISSUE þ a11ABSPMDA þ
YearFE þ IndustryFE (1)
The dependent variable FirmValue is proxied by Tobin's
Q, calculated as the sum of the market value of equity and
book value of total liabilities, divided by the book value of
total assets. Prior research uses Tobin's Q primarily to mea-
sure a firm's value (e.g., Cahan et al., 2016). To verify the
robustness of the results, we use both Tobin's Q (TobinQ) and
adjusted Tobin's Q (ADJ_TobinQ). We calculate the adjusted
Tobin's Q by subtracting the mean Tobin's Q value of each
industry-year from the Tobin's Q of each observation. To
clarify the causal relationship and to exclude the results from
the possible inverse relationship between CSR and firm
value, we also use the one-year-ahead Tobin's Q (TobinQtþ1,
ADJ_TobinQtþ1).
The variable of interest is CSR. We use both internal and
external CSR (CSRInternal, CSRExternal) in the regression. As
mentioned above, we use the WWGI index, which assigns a
3
Therefore, firms classified as “Good” receive a value of 3.

W.J. Lee, S.U. Choi
value of 1e5 based on the firms’ contribution to suppliers to
proxy for internal CSR. We use the CSR score from the
KEJI as a proxy of external CSR. If conducting CSR ac-
tivities increases firm value, we expect a positive coefficient
on CSR (a1 > 0). We run the regression separately for each
CSR activity because the CSR indexes are distributed by
different institutions. Thus, we conduct within-sample
analyses.
We add control variables based on prior research (Kalay &
Shimrat, 1987; Harjoto & Jo, 2015). First, we include the
natural logarithm of total assets (SIZE ) to control for firm
size. Financial structure also affects firm value; we therefore
introduce the leverage ratio (LEV) of total liabilities to total
assets to control for the financial structure. To control for
liquidity, we add the current ratio (CURR), representing
current assets divided by current liabilities. We use the sum
of inventories and accounts receivables to total assets
(INVREC ) to control for the firm's complexity. Firm perfor-
mance is a direct determinant of firm value, so we control for
operating cash flow divided by total assets (CF ) and for
operating performance (ROA), which we calculate as net in-
come divided by total assets. We also include firms reporting
losses (LOSS ). Additionally, varying growth opportunities
affect firm value, so we include sales growth (SalesGRW ). A
large stream of research documents a price drop after stock
issues (Kalay & Shimrat, 1987). Hence, we add a dummy
variable, ISSUE, equal to one if a firm issues a number of
shares equal to or greater than 10 percent of its previous
number of shares, and zero otherwise. Previous studies
generally indicate that earnings management distorts capital
market resource allocation, reducing firm value. We thereby
include discretionary accruals, specifically the absolute value
of performance-matched discretionary accruals (ABSPMDA),
to control for the effect of earnings manipulation on firm
value. Lastly, we include year (Year FE ) and industry fixed
effects (Industry FE ) to control for the time-series changes
and industry-specific characteristics of dependent variables.
All t-values reported in the tables are clustered at the firm
level.
Next, we apply Eq. (2) to test H3. Here, we integrate both
indexes in one regression model to test the combined effect of
internal and external CSR on firm value. We do this by con-
structing indicator variables for each CSR score if included in
each index. For instance, in Eq. (2), CSRInt&Ext indicates ob-
servations included in both indexes. CSRInt_Only indicates firms
included only in the WWGI index and CSRExt_Only indicates
those included only in the KEJI index. As we construct these
variables as dummy variables, the intercept captures the effect
of no CSR engagement on firm value.
Table 1
Yearly distribution.
Year 2013 2014 2015 2016 2017 2018
Full sample 1435 1495 1565 1640 1720 1806
Internal CSR sample 64 71 84 94 113 119
External CSR sample 168 170 165 170 161 166
+ MODEL
_
Borsa Istanbul Review xxx (xxxx) xxx
FirmValue ¼ a0 þ a1CSRInt&Ext þ a2CSRExt_Onlyþ
a3CSRInt_Only þ Controls þ YearFE þ
IndustryFE (2)
In our interpretation, a firm investing in both types of CSR
shows increased value if the coefficient of CSRInt&Ext (a1) is
positively significant. The coefficients of CSRExt_Only and
CSRInt_Only indicate the influence on firm value for firms
investing in external or internal CSR only, respectively. Eq. (2)
contains the same controls as in Eq. (1) and the t-values are
clustered at the firm level.
3.4. The sample
The sample consists of firms listed in the Korean stock
market. Korea adopted IFRS in 2011 to align its accounting
standard with the highest international standard. We restrict
our sample to December year-end firms to maintain compa-
rability of the financial data among firms. We also use firms
operating in non-financial industries to reduce the effect of
different types of financial statements. Finally, we use the
firm-year observations that have data for all variables for each
test, resulting in a final sample of 9661 observations.
Table 1 shows the sample distribution by year. Our sample
starts in 2013, as it is the most recent year that uses the new
evaluation method to generate the KEJI score. The year 2013
has 1435 observations based on the full sample criteria.
Among the 1435 observations, 64 and 168 firms have a WWGI
score (internal CSR index) and a KEJI score (external CSR
index), respectively.4 There is an increasing trend in the in-
ternal CSR sample, while the external CSR sample is
distributed within the range of 160e170. This distribution
arises because the KEJI discloses 200 firms every year, but the
number of companies included in the WWGI increases over
time. We lose observations due to missing financial data. The
total number of observations with WWGI and KEJI scores are
545 and 1000, respectively.
4. Empirical results
4.1. Descriptive statistics
Panels A and B of Table 2 present the descriptive statistics
and Pearson correlations. In Panel A, the mean of CSRInternal is
3.618. This variable has a value from 1 to 5, which matches the
assigned grades from lowest to highest. The mean value of 3.618
therefore shows that a relatively large number of firms are
distributed in the higher grade. The mean of CSRExternal is
63.759. The mean values of CSRInt&Ext, CSRExt_Only, and
CSRInt_Only are 0.012, 0.092, and 0.044, respectively. The sum of
the mean values of the three variables is 0.148, indicating that
14.8 percent of firms are involved in either internal, external, or
Total
9661
545 4
The internal and external CSR firms are not mutually exclusive samples;
1000
some firms appear in both indexes.
6
 Table 2

W.J. Lee, S.U. Choi

Descriptive statistics and correlation matrix.
Panel A. Descriptive statistics
Variables Mean SD P1 Q1 Median Q3 P99
CSRInternal 3.618 0.970 1.000 1.000 3.000 4.000 5.000
CSRExternal 63.759 2.027 59.330 62.405 63.535 64.970 69.325
CSRInt&Ext 0.012 0.109 0.000 0.000 0.000 0.000 1.000
CSRExt_Only 0.092 0.288 0.000 0.000 0.000 0.000 1.000
CSRInt_Only 0.044 0.206 0.000 0.000 0.000 0.000 1.000
TobinQ 1.497 1.114 0.489 0.868 1.132 1.663 7.289
ADJ_TobinQ 0.031 0.988 2.254 0.503 0.195 0.162 4.939
TobinQtþ1 1.504 1.123 0.491 0.871 1.131 1.671 7.289
ADJ_TobinQtþ1 0.031 1.008 2.191 0.510 0.205 0.162 5.065
Asset2TW 0.061 0.240 0.000 0.000 0.000 0.000 1.000
Group 0.426 0.495 0.000 0.000 0.000 1.000 1.000
Manufacturer 0.671 0.470 0.000 0.000 1.000 1.000 1.000
SIZE 18.988 1.348 16.623 18.065 18.737 19.630 23.713
LEV 0.372 0.202 0.026 0.205 0.366 0.520 0.873
CURR 3.208 4.967 0.207 1.008 1.672 3.194 36.584
INVREC 0.249 0.158 0.000 0.130 0.231 0.345 0.712
CF 0.039 0.084 0.250 0.002 0.041 0.086 0.258
ROA 0.012 0.116 0.479 0.013 0.025 0.064 0.340
LOSS 0.290 0.454 0.000 0.000 0.000 1.000 1.000
SalesGRW 0.063 0.348 0.677 0.087 0.021 0.135 1.991

+
ISSUE 0.174 0.379 0.000 0.000 0.000 0.000 1.000

MODEL
7

ABSPMDA 0.062 0.066 0.001 0.018 0.042 0.081 0.373

Panel B: Correlation matrix
Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
(1)CSRInternal 1.000
(2)CSRExternal 0.093 1.000
(3)TobinQ 0.144 0.188 1.000
(4)ADJ_TobinQ 0.102 0.059 0.824 1.000
(5)TobinQtþ1 0.130 0.193 0.793 0.632 1.000
(6)ADJ_TobinQtþ1 0.079 0.027 0.624 0.746 0.834 1.000
(7)Asset2TW 0.483 0.080 ¡0.044 0.007 ¡0.050 0.016 1.000
(8)Group 0.260 0.019 ¡0.082 ¡0.050 ¡0.083 ¡0.052 0.261 1.000
(9)Manufacturer ¡0.103 0.115 ¡0.095 0.007 ¡0.092 0.008 ¡0.063 ¡0.169 1.000
(10)SIZE 0.595 0.164 ¡0.221 ¡0.135 ¡0.246 ¡0.159 0.669 0.410 ¡0.042 1.000

Borsa Istanbul
(11)LEV ¡0.153 ¡0.080 ¡0.108 ¡0.021 ¡0.089 0.007 0.093 0.015 0.087 0.158 1.000
(12)CURR 0.022 0.014 0.153 0.063 0.130 0.051 ¡0.077 ¡0.053 ¡0.095 ¡0.134 ¡0.519 1.000

_
(13)INVREC ¡0.320 0.031 ¡0.087 ¡0.048 ¡0.063 ¡0.037 ¡0.126 ¡0.157 0.310 ¡0.131 0.293 ¡0.213 1.000
(14)CF 0.191 0.058 ¡0.072 ¡0.024 ¡0.077 ¡0.031 0.058 0.062 0.057 0.173 ¡0.143 ¡0.021 ¡0.058 1.000
(15)ROA 0.196 0.132 ¡0.095 ¡0.048 ¡0.100 ¡0.059 0.048 0.083 0.039 0.207 ¡0.277 0.064 0.029 0.545 1.000

Review xxx (xxxx) xxx

(16)LOSS ¡0.093 0.015 0.105 0.059 0.118 0.077 ¡0.051 ¡0.084 0.006 ¡0.197 0.253 ¡0.051 ¡0.036 ¡0.427 ¡0.690 1.000
(17)SalesGRW 0.013 0.126 0.150 0.108 0.100 0.067 0.005 0.014 ¡0.044 0.007 0.001 0.011 0.026 0.081 0.180 ¡0.156 1.000
(18)ISSUE 0.050 0.025 0.139 0.103 0.129 0.105 ¡0.060 ¡0.078 0.008 ¡0.168 0.058 0.020 ¡0.032 ¡0.181 ¡0.188 0.168 0.094 1.000
(19)ABSPMDA 0.016 0.010 0.187 0.135 0.148 0.127 ¡0.079 ¡0.071 ¡0.076 ¡0.168 0.055 0.014 0.027 ¡0.183 ¡0.160 0.153 0.059 0.198 1.000
Panel A: Descriptive statistics are based on 9661 observations except for CSRInternal (545 observations), CSRExternal (1000 observations), and TobinQtþ1, ADJ_TobinQtþ1 (7839 observations).
Panel B: Bold indicates significant at 5% or higher level.
 + MODEL
W.J. Lee, S.U. Choi
both internal and external CSR activities. The mean values of the
four dependent variables are 1.497, 0.031, 1.504, and 0.031
for TobinQ, ADJ_TobinQ, TobinQtþ1, and ADJ_TobinQtþ1,
respectively. For the instrument variables, the mean value of
Asset2TW is 0.061, indicating that 6.1 percent of companies report
total assets of more than 2 trillion won. The mean of Group is
0.426, showing that 42.6 percent of firms belong to a business
group. The mean of Manufacturer is 0.671, indicating that 67.1
percent of firms operate in manufacturing industries.
For the control variables, the mean values of SIZE, LEV,
CURR, and INVREC are 18.988, 0.372, 3.208, and 0.249,
respectively. The mean values of CF, ROA, LOSS, and
SalesGRW are 0.039, 0.012, 0.290, and 0.063, respectively.
Approximately 17.4 percent of the companies issued a number
of stocks equal to or more than 10 percent of its previous
outstanding shares (ISSUE ). Lastly, the mean value of
ABSPMDA is 0.062. All the statistics are similar to the values
reported in prior research using Korean listed firms.
Panel B of Table 2 reports the Pearson correlations for
selected variables. CSRInternal is positively correlated with
three out of the four proxies of firm value. CSRExternal is
positively related with TobinQ and TobinQtþ1. Overall, this
evidence suggests that CSR activities generally enhance firm
value, at least in univariate correlations. The three instruments
are significantly correlated with CSR variables, indicating that
they explain the CSR activities adequately.
4.2. Test results of H1 and H2 (within-sample analyses)
Tables 3 and 4 show the results of the within-sample ana-
lyses. Since different institutions provide the two CSR in-
dexes, we conduct the analyses using the sample of
observations included in each CSR index.
First, Table 3 reports the within-sample analyses using the
WWGI scores from the KCCP. It presents the test results of the
multivariate regression analyses of H1. In model (1), we report
the results using TobinQ as the dependent variable. Consistent
with H1, we find a positive coefficient of CSRInternal, with a
value of 0.180 and t-value of 2.14, significant at the 5 percent
level. The coefficients of CSRInternal are significantly positive
for the three other proxies of firm value. For instance, the
coefficients of CSRInternal are 0.195, 0.173, and 0.179 for the
ADJ_TobinQ, TobinQtþ1, and ADJ_TobinQtþ1 models,
respectively. To summarize, internal CSR activities for the
business partners in a supply chain increase firm value.
Particularly, the results are consistent when using the current
and one-year-ahead proxies of firm value. The results are also
significant for both Tobin's Q and industry-adjusted Tobin's Q.
The results for the control variables are similar to the ex-
ante predictions. First, firm size (SIZE ) is negatively related
with Tobin's Q. Since Tobin's Q is the market value scaled by
total assets, it is mechanically negatively related to total assets.
CF and ROA show positive and significant coefficients, indi-
cating high market values for firms reporting higher operating
performance. ISSUE has negative coefficients, consistent with
the results of prior studies showing that the performance of
firms issuing stocks declines afterwards.
8
_
Borsa Istanbul Review xxx (xxxx) xxx
Next, Table 4 reports the within-sample analyses using
external CSR scores from the KEJI to test H2. Similar to Table
3, we report the results using TobinQ as the dependent variable
in model (1). Consistent with H2, we find a positive coefficient
of CSRExternal (coefficient ¼ 0.053, t-value ¼ 3.44) significant
at the 1 percent level. We also find that CSRExternal is positively
significant using three other proxies of firm value. The co-
efficients of CSRExternal are 0.046, 0.057, and 0.051 for the
ADJ_TobinQ, TobinQtþ1, and ADJ_TobinQtþ1 models,
respectively (t-values are 3.01, 3.75, and 3.26, respectively). In
summary, the CSR activities for external stakeholders signif-
icantly enhance firm value.
Overall, the findings support H1 and H2, suggesting that CSR
firms have higher firm value. Our test results expand the litera-
ture by demonstrating that CSR activities targeting not only
external stakeholders but also internal suppliers contribute to
firm value.
4.3. Test results of H3 (full sample analyses)
Our analyses in the previous tables are based on a within-
sample regression because the CSR indexes are provided by
two different institutions. Here, we integrate both indexes in
one regression model by constructing indicator variables for
each CSR type for firms that engage in both types of CSR
activity. For instance, CSRInt&Ext indicates observations that
appear in both indexes. In contrast, CSRInt_Only indicates firms
included only in the WWGI, while CSRExt_Only indicates
companies only in the KEJI index.
Table 5 presents the results. The coefficients of CSRInt&Ext
are positive and significant in all columns (coefficients ¼ 0.517,
0.481, 0.640, 0.609, t-values ¼ 4.54, 4.18, 4.43, 4.19, respec-
tively). The coefficients of CSRExt_Only are negative and sig-
nificant, while those of CSRInt_Only are positive and significant.
We interpret these findings as follows. First, firms conducting
both CSR types have higher firm values. Second, companies
conducting only external CSR have lower values than those that
are not engaged in any CSR activities. Third, firms conducting
only internal CSR have higher values. Fourth, to summarize
these results, the positive relationship between CSR activity and
firm value derives from internal CSR.
This is a new evidence in the related literature. One stream
of previous studies reports that external CSR activities are
beneficial to firm value, while other studies argue that these
activities harm firm value. Our results show researchers should
consider internal CSR as an additional factor affecting the
relationship between CSR and firm value. Firms that lack
appropriate social contribution activities for their internal
suppliers within a supply chain cannot efficiently reflect their
external CSR efforts on their value.
4.4. Additional analyses to mitigate selection bias and
endogeneity
We can expect a possible endogenous relationship between
CSR and firm value. Thus, it is essential to mitigate any
endogeneity. The first concern is that the sample in each CSR
 + MODEL

W.J. Lee, S.U. Choi _

Borsa Istanbul Review xxx (xxxx) xxx

Table 3
Internal CSR and firm value.
Dependent variable ¼
Variables (1) TobinQ (2) ADJ_TobinQ (3) TobinQtþ1 (4) ADJ_TobinQtþ1
Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value)
Intercept 5.450*** 4.194*** 5.644*** 4.227**
(3.49) (2.63) (3.17) (2.43)
CSRInternal 0.180** 0.195** 0.173* 0.179*
(2.14) (2.29) (1.74) (1.89)
SIZE 0.247*** 0.255*** 0.252*** 0.245***
(3.08) (3.12) (2.64) (2.65)
LEV 0.403 0.355 0.386 0.298
(1.01) (0.85) (0.76) (0.60)
CURR 0.035 0.036 0.024 0.031
(0.38) (0.38) (0.28) (0.35)
INVREC 0.143 0.105 0.387 0.300
(0.27) (0.20) (0.57) (0.44)
CF 3.081*** 3.283*** 2.647*** 2.668***
(3.98) (4.08) (3.31) (3.27)
ROA 5.961*** 5.662*** 4.958*** 4.515**
(3.41) (3.16) (2.65) (2.39)
LOSS 0.296** 0.284** 0.263** 0.231*
(2.34) (2.17) (2.01) (1.76)
SalesGRW 0.059 0.075 0.030 0.044
(0.33) (0.37) (0.21) (0.31)
ISSUE 0.254** 0.242** 0.200* 0.168
(2.31) (2.15) (1.66) (1.41)
ABSPMDA 0.485 0.547 0.324 0.375
(0.54) (0.55) (0.32) (0.36)
Year Included Included Included Included
Indsutry Included Included Included Included
ADJ. R2 0.473 0.395 0.466 0.390
N of obs. 545 545 426 426
T-values are clustered at the firm-level. ***, **, and * denote significance at 1, 5, and 10% level, respectively. We lose 119 observations in model (3) and (4) as they
use one-year-ahead firm value proxies.
Bold indicates the variable of interest.

index might suffer a self-selection bias. The firms selected by
each institution may have different characteristics than those not
selected, and these characteristics may influence the results. To
address this issue, we run traditional two-stage regressions. In
Eq. (3), we generate the Inverse Mills Ratio (IMR), which we
then include in Eq. (4). This kind of two-stage regression is
widely used to mitigate existing bias in a self-selected sample.
CSR ¼ b0 þ b1SIZE þ b2LEV þ b3CURR þ b4INVREC þ
b5CF þ b6ROA þ b7LOSS þ b8SalesGRW
þ b9ISSUE þ b10ABSPMDA þ YearFE
þ IndustryFE (3)
Firm Value ¼ a0 þ a1CSR þ Controls þ IMR þ YearFE þ
IndustryFE (4)
We include the following determinants of CSR in Eq. (3).5
First, we add SIZE to control for the firm size effect and other
omitted firm characteristics. We also use LEV to control the
financial structure effect on CSR. CURR, INVREC, and CF
5
We include the same controls in Eq. (1) in the first-stage regression (Eq.
(3)) to avoid model misspecification (Lennox, Francis, & Wang, 2012).
control for liquidity and a firm's operating complexity.
McGuire, Sundgren, and Schneeweis (1988) find that ac-
counting performance is associated with CSR activities;
therefore, we also include ROA and LOSS. Because a firm's
growth rate is related to CSR activities, we include SalesGRW.
Further, to control for CSR activities in firms issuing new
stocks, we include ISSUE. An increasing number of CSR
studies investigate the relationship between accounting trans-
parency and CSR. For instance, Kim, Park, and Wier (2012)
show that firms involved in CSR are negatively associated
with discretionary accruals. Thus, we include ABSPMDA.
Although untabulated, the results of the first-stage probit
regression show that the SIZE, INVREC, and CF are positively
related with the dependent variable of WWGI, where WWGI
equals to one if a firm is included in the KCCP's disclosures. In
contrast, LEV, LOSS, SalesGRW, ISSUE, and ABSPMDA are
negatively related with WWGI. When KEJI is used as the
dependent variable, which is equal to one if a firm is included
in the KEJI index, the independent variables are qualitatively
similar to those of WWGI model in their direction.
Panel A of Table 6 presents the second-stage regression
results for internal and external CSR firms after including
IMR. The coefficients of CSRInternal are positive and significant
in columns (1) and (2) (coefficients ¼ 0.199, 0.188, t-

9
 + MODEL

W.J. Lee, S.U. Choi _

Borsa Istanbul Review xxx (xxxx) xxx

Table 4
External CSR and firm value.
Dependent variable ¼
Variables (1) TobinQ (2) ADJ_TobinQ (3) TobinQtþ1 (4) ADJ_TobinQtþ1
Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value)
Intercept 1.515 2.312** 1.370 2.519**
(1.38) (2.09) (1.20) (2.14)
CSRExternal 0.053*** 0.046*** 0.057*** 0.051***
(3.44) (3.01) (3.75) (3.26)
SIZE 0.063** 0.073** 0.083*** 0.084***
(2.10) (2.41) (2.82) (2.87)
LEV 0.728*** 0.701*** 0.515** 0.462*
(3.33) (3.22) (1.97) (1.78)
CURR 0.002 0.002 0.007 0.007
(0.37) (0.39) (1.10) (1.05)
INVREC 0.571* 0.715** 0.680** 0.732**
(1.69) (2.11) (2.11) (2.25)
CF 2.215*** 2.153*** 2.473*** 2.432***
(3.45) (3.44) (3.09) (2.82)
ROA 2.812*** 2.810*** 2.539* 2.445*
(2.64) (2.66) (1.77) (1.65)
LOSS 0.093 0.072 0.039 0.030
(0.88) (0.67) (0.32) (0.24)
SalesGRW 0.278** 0.321*** 0.285 0.335
(2.20) (2.62) (1.36) (1.57)
ISSUE 0.189* 0.161 0.326*** 0.345***
(1.91) (1.55) (2.66) (2.70)
ABSPMDA 0.682 0.490 1.726 2.033*
(0.79) (0.57) (1.56) (1.72)
Year Included Included Included Included
Indsutry Included Included Included Included
ADJ. R2 0.258 0.326 0.244 0.308
N of obs. 1000 1000 832 832
T-values are clustered at the firm-level. ***, **, and * denote significance at 1, 5, and 10% level, respectively. We lose 168 observations in model (3) and (4) as they
use one-year-ahead firm value proxies.
Bold indicates the variable of interest.

value ¼ 2.27, 1.91, respectively). Columns (3) and (4) show
the second-stage regressions results for external CSR firms.
After controlling for IMR, all models show that external CSR
activities increase firm value.6 Overall, the two-stage regres-
sion test results confirm that our results are robust when
including IMR.
Our next tests aim to address the concern that the results
might suffer from a potential endogenous relationship between
CSR and firm value. In contrast with the previous self-
selection bias, this does not originate from the different CSR
indexes, but from the inherent characteristics of CSR and firm
value. We mitigate this concern by using an instrumental
variables approach.
In the first-stage regression, we use three instrument
variables systematically related with CSR activities. First,
the Asset2TW variable equals one if a firm's total asset value is
greater than 2 trillion Korean won,7 as firms reporting more
than 2 trillion won of total assets are subject to different
6 relationship with suppliers compared with stand-alone firms.
Although untabulated for brevity, we find positive and significant co-
efficients of CSR variables in the TobinQ and TobinQtþ1 models when we run
the analyses using the same independent variables shown in Table 6.
7
The equivalent of 2 trillion Korean won is about 1.7 billion US dollars.
10
regulations than companies in a lower category. For example,
these large firms are required to hire at least three outside
directors, and the ratio of the number of outside directors to
the total number of directors should exceed 50 percent.
Furthermore, they must have an audit committee in their
boards. These characteristics indicate that these firms may
have completely different corporate governance mechanisms
as do small firms, which might affect the firms' CSR activ-
ities. Since this stronger regulation of large firms is an
exogenous variable, and is thus not necessarily related to
firm value, prior research using the Korean stock market
generally uses this 2 trillion won restriction as an instrument
variable (e.g., Black, Jang, & Kim, 2006). Second, we use
the Group variable, which equals one if a firm belongs to a
business group. Generally, such companies are highly likely
to be linked to secondary and tertiary suppliers in their
supply chains because a business group is a conglomerate
entity consisting of several related firms. Therefore, firms
that belong to such groups will pay more attention to their
In particular, the KCCP publishes WWGI scores for firms in
business groups with a great social impact. Third, we include
Manufacturer variable, which equals one for manufacturing
 + MODEL

W.J. Lee, S.U. Choi _

Borsa Istanbul Review xxx (xxxx) xxx

Table 5
Internal and external CSR and firm value.
Dependent variable ¼
Variables (1) TobinQ (2) ADJ_TobinQ (3) TobinQtþ1 (4) ADJ_TobinQtþ1
Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value)
Intercept 4.923*** 2.335*** 5.612*** 2.902***
(12.41) (5.77) (12.92) (6.57)
CSRInt&Ext 0.517*** 0.481*** 0.640*** 0.609***
(4.54) (4.18) (4.43) (4.19)
CSRExt_Only ¡0.114** ¡0.130*** ¡0.110** ¡0.131**
(2.34) (2.65) (2.04) (2.40)
CSRInt_Only 0.469*** 0.471*** 0.537*** 0.512***
(5.04) (5.06) (5.61) (5.32)
SIZE 0.152*** 0.151*** 0.189*** 0.181***
(7.81) (7.68) (8.78) (8.29)
LEV 0.094 0.041 0.176 0.137
(0.86) (0.38) (1.45) (1.11)
CURR 0.017** 0.010 0.014* 0.009
(2.39) (1.44) (1.67) (1.05)
INVREC 0.424*** 0.415*** 0.391** 0.381**
(2.88) (2.81) (2.43) (2.35)
CF 0.240 0.444** 0.221 0.441**
(1.13) (2.17) (1.01) (2.06)
ROA 0.174 0.033 0.041 0.072
(0.73) (0.14) (0.17) (0.31)
LOSS 0.092** 0.083** 0.131*** 0.119***
(2.34) (2.25) (3.01) (2.86)
SalesGRW 0.381*** 0.317*** 0.249*** 0.203***
(9.55) (7.75) (6.18) (4.78)
ISSUE 0.135*** 0.133*** 0.139*** 0.145***
(4.26) (4.14) (4.13) (4.28)
ABSPMDA 1.769*** 1.639*** 1.294*** 1.399***
(7.84) (7.34) (5.04) (5.40)
Year Included Included Included Included
Indsutry Included Included Included Included
ADJ. R2 0.262 0.072 0.249 0.069
N of obs. 9661 9661 7839 7839
T-values are clustered at the firm-level. ***, **, and * denote significance at 1, 5, and 10% level, respectively. We lose 1822 observations in model (3) and (4) as
they use one-year-ahead firm value proxies.
Bold indicates the variable of interest.

firms because the relationship between a manufacturing 4.5. Additional tests

company and its partners is more important than that in other
industries. We use CSRInt&Ext as a dependent variable and
include the three abovementioned instruments and control
variables as independent variables. We then estimate the
fitted value of the dependent variable (Pred_CSRInt&Ext).
In Panel B of Table 6, the column (1) shows that Asset2TW
and Manufacturer are positive and significant among the three
instruments. Although untabulated, we find significant statis-
tics for the weak instrument test, suggesting that our instru-
ment variables are not weakly identified. Columns (2) and (3)
present the results using the fitted value of CSR on ADJ_To-
binQ and ADJ_TobinQtþ1, respectively. The coefficients of
Pred_CSRInt&Ext are 6.078 and 6.204 (t-values ¼ 8.53, 8.37),
respectively. Thus, our results are robust when using an
instrumental approach, suggesting that an endogenous rela-
tionship does not drive our findings.
We next examine whether promotional activities related to
business partners within a supply chain influence our results.
A donation to suppliers and customers in a supply chain may
speed up the conversion of inventory assets into cash. For
instance, firms can donate to their business partner to shorten
the manufacturing, delivery, and payment time. These effects
are indicated by accounting-based numbers such as inventory
days (the days it takes to sell the inventory) and accounts
receivable days (the days it takes to clear the accounts re-
ceivable). A decrease in both of these areas indicates the
timeliness of the operating cycle (Lo, Yeung, & Cheng,
2009). Therefore, the promotion provided to business part-
ners in the supply chain will speed up the firm's cash con-
version rate, which can be proxied by these indicators. We
therefore use the internal CSR indicator as a dependent

11
 + MODEL

W.J. Lee, S.U. Choi _

Borsa Istanbul Review xxx (xxxx) xxx

Table 6
Applying two-stage regressions and instrumental approach to mitigate endogeneity.
Panel A. Two-stage regressions
Variables Dependent variable ¼
(1) ADJ_TobinQ (2) ADJ_TobinQtþ1 (3) ADJ_TobinQ (4) ADJ_TobinQtþ1
Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value)
Intercept 3.219* 2.261 2.119 5.526*
(1.79) (1.23) (0.73) (1.89)
CSRInternal 0.199** 0.188*
(2.27) (1.91)
CSRExternal 0.046*** 0.051***
(3.01) (3.27)
IMR 0.143 0.285 0.035 0.541
(0.62) (1.01) (0.07) (1.13)
Controls&Year&Industry Included Included Included Included
ADJ. R2 0.395 0.394 0.325 0.309
N of obs. 545 426 1000 832
Panel B. Instrumental approach
Variables 1st stage 2nd stage
(1) Dep ¼ CSRInt&Ext (2) Dep ¼ ADJ_TobinQ (3) Dep ¼ ADJ_TobinQtþ1
Coefficient (T-Value) Coefficient (T-Value) Coefficient (T-Value)
Intercept 0.169*** 3.525*** 4.007***
(7.35) (11.21) (11.87)
Asset2TW 0.085***
(14.09)
Manufacturer 0.014***
(5.79)
Group 0.002
(0.86)
Pred_CSRInt&Ext 6.078*** 6.204***
(8.53) (8.37)
Controls&Year Included Included Included
ADJ. R2 0.078 0.043 0.043
N of obs. 9661 9661 7839
Control variables are included but omitted for brevity. In Panel B, industry fixed effects are not included due to possible collinearity with Manufacturer variable. T-
values are clustered at the firm-level. ***, **, and * denote significance at 1, 5, and 10% level, respectively.
Bold indicates the variable of interest.

variable and inventory days, account receivable days, and
other controls as independent variables and estimate the re-
sidual value from the model. We use the residuals as a proxy
for internal CSR activity after controlling for promotional
donations to business partners within the supply chain. We
then test whether the residuals are positively related to firm
value. Untabulated results show that the coefficients of the
residuals are positive and significant at the 5 percent level or
above. Overall, after controlling for the promotional activ-
ities from internal CSR, the remaining internal CSR still
improves firm value.
Lastly, the purpose of disclosing the KEJI index is to induce
sustainable development of social enterprises and establish
social contribution activities. As the index is provided by an
institute, its users have less discretion when selecting the ac-
tivities and points. In the score, 15 points are allocated to the
social contribution subcategory. We re-test our hypothesis by
regenerating the points of this subcategory, increasing them
from 15 to 30, to give this category the largest weight. Thus,
the regenerated external CSR score consists of six categories
with 115 points in total. We then rerun the analysis reported in
Table 4 using the new score. Although untabulated, we find a
positive coefficient of CSRExternal for all four different TobinQ
proxies. Thus, we find that our results are robust in different
scoring methods.
5. Conclusion
This study investigates whether CSR practices related to
suppliers, which affect the objective quality of the product,
and those targeting external stakeholders, which shape sub-
jective perceptions, will positively affect a firm's value. By
dividing CSR into internal and external activities, we aim to
broaden the scope of CSR-related research.
We analyze firms listed in the Korean stock market using
an internal CSR index from the KCCP and an external one
from the KEJI and find that both types of CSR lead to higher
firm value. Furthermore, a core contribution of our study is
the finding that the positive relation between CSR and firm
value derives mainly from internal CSR activities. The

12
 + MODEL
W.J. Lee, S.U. Choi
results show that firms involved in both types of CSR have
higher values, while those engaged only in external CSR
have lower values than non-CSR firms. Interestingly, we find
that firms investing only in internal CSR also show higher
values than non-CSR firms, indicating that social activity for
internal suppliers is a core CSR aspect that determines firm
value.
These findings are meaningful because they provide a
specific path indicating how social activities increase firm
value, offering clearer, more direct results than merely
claiming that external social contributions are vaguely related
to a firm's value. Also, existing research provides limited
analysis of whether CSR activities targeting suppliers have
positive effects on a firm's value. The literature contains
limited empirical evidence using few observations. Our study
contributes to the literature by expanding the concept of CSR
to a broader scale.
Our study also has several caveats. First, the two CSR
proxies are published by different institutions. Therefore, there
are inherent differences in the calculation of each index.
Second, our findings are based on the Korean market, and thus
are not generalizable to other jurisdictions. Future research
may report further evidence from other emerging or developed
markets. Finally, we acknowledge that although we apply
several methods to the analyses to mitigate endogeneity con-
cerns, we cannot argue that the model completely controlled
for them. Therefore, it is necessary to be cautious with the
interpretation of our results.
Declaration of competing interest
None.
References
Bhandari, A., & Javakhadze, D. (2017). Corporate social responsibility and
capital allocation efficiency. Journal of Corporate Finance, 43, 354e377.
Bhardwaj, P., Chatterjee, P., Demir, K. D., & Turut, O. (2018). When and how
is corporate social responsibility profitable? Journal of Business Research,
84, 206e219.
Black, B. S., Jang, H., & Kim, W. (2006). Does corporate governance predict
firms' market values? Evidence from Korea. Journal of Law, Economics,
and Organization, 22(2), 366e413.
Bosse, D. A., & Coughlan, R. (2016). Stakeholder relationship bonds. Journal
of Management Studies, 53, 1197e1222.
Buchanan, B., Cao, C. X., & Chen, C. (2018). Corporate social responsibility,
firm value, and influential institutional ownership. Journal of Corporate
Finance, 52, 73e95.
Cahan, S. F., De Villiers, C., Jeter, D. C., Naiker, V., & Van Staden, C. J.
(2016). Are CSR disclosures value relevant? Cross-country evidence.
European Accounting Review, 25(3), 579e611.
Carroll, A. (1979). A three dimensional model of corporate performance.
Academy of Management Review, 4, 497e505.
Castelo, M., & Lima, L. (2006). Corporate social responsibility and resource-
based perspectives. Journal of Business Ethics, 69, 111e132.
Chang, Y. K., Oh, W. Y., Park, J. H., & Jang, M. G. (2017). Exploring the
relationship between board characteristics and CSR: Empirical evidence
from Korea. Journal of Business Ethics, 140(2), 225e242.
Cheng, B., Ioannou, I., & Serafeim, G. (2014). Corporate social responsibility
and access to finance. Strategic Management Journal, 35, 1e23.
13
_
Borsa Istanbul Review xxx (xxxx) xxx
Du, X. (2015). Is corporate philanthropy used as environmental misconduct
dressing? Evidence from Chinese family-owned firms. Journal of Business
Ethics, 129, 341e361.
Feng, Y., Zhu, Q., & Lai, K. H. (2017). Corporate social responsibility for
supply chain management: A literature review and bibliometric analysis.
Journal of Cleaner Production, 158(August), 296e307.

Gallego-Alvarez, I., Prado-Lorenzo, J. M., & García-Sanchez, I.-M. (2011).
Corporate social responsibility and innovation: A resource-based theory.
Management Decision, 49(10), 1709e1727.
Harjoto, M. A., & Jo, H. (2015). Legal vs. normative CSR: Differential impact
on analyst dispersion, stock return volatility, cost of capital, and firm value.
Journal of Business Ethics, 128(1), 1e20.
Hsueh, C. F. (2014). Improving corporate social responsibility in a supply
chain through a new revenue sharing contract. International Journal of
Production Economics, 151(May), 214e222.
Kalay, A., & Shimrat, A. (1987). Firm value and seasoned equity issues: Price
pressure, wealth redistribution, or negative information. Journal of
Financial Economics, 19(1), 109e126.
Kim, Y., Park, M. S., & Wier, B. (2012). Is earnings quality associated with
corporate social responsibility? The Accounting Review, 87(3), 761e796.
Kwon, I. W. G., & Suh, T. (2005). Trust, commitment and relationships in
supply chain management: A path analysis. Supply Chain Management:
International Journal, 10(1), 26e33.
Lee, H. Y., Kwak, D. W., & Park, J. Y. (2017). Corporate social responsibility
in supply chains of small and medium-sized enterprises. Corporate Social
Responsibility and Environmental Management, 24(6), 634e647.
Lennox, C. S., Francis, J. R., & Wang, Z. (2012). Selection models in ac-
counting research. The Accounting Review, 87(2), 589e616.
Lenz, I., Wetzel, H. A., & Hammerschmidt, M. (2017). Can doing good lead to
doing poorly? Firm value implications of CSR in the face of CSI. Journal
of the Academy of Marketing Science, 45(5), 677e697.
Lindgreen, A., Swaen, V., Maon, F., Andersen, M., & Skjoett-Larsen, T.
(2009). Corporate social responsibility in global supply chains. Supply
Chain Management: International Journal, 14(2), 75e86.
Lo, C. K., Yeung, A. C., & Cheng, T. C. E. (2009). ISO 9000 and supply chain
efficiency: Empirical evidence on inventory and account receivable days.
International Journal of Production Economics, 118(2), 367e374.
Maignan, I., Ferrell, O. C., & Hult, G. T. M. (1999). Corporate citizenship:
Cultural antecedents and business benefits. Journal of the Academy of
Marketing Science, 27(4), 455e469.
Maloni, M. J., & Brown, M. E. (2006). Corporate social responsibility in the
supply chain: An application in the food industry. Journal of Business
Ethics, 68(1), 35e52.
Manchiraju, H., & Rajgopal, S. (2017). Does corporate social responsibility
(CSR) create shareholder value? Evidence from the Indian companies Act
2013. Journal of Accounting Research, 55(5), 1257e1300.
Marquis, C., & Qian, C. (2014). Corporate social responsibility reporting in
China: Symbol or substance? Organization Science, 25, 127e148.
McGuire, J. B., Sundgren, A., & Schneeweis, T. (1988). Corporate social re-
sponsibility and firm financial performance. Academy of Management
Journal, 31(4), 854e872.
Nekhili, M., Nagati, H., Chtioui, T., & Rebolledo, C. (2017). Corporate social
responsibility disclosure and market value: Family versus nonfamily firms.
Journal of Business Research, 77, 41e52.
Park, H. J., & Ha, M. H. (2020). Corporate social responsibility and earnings
transparency: Evidence from Korea. Corporate Social Responsibility and
Environmental Management, 27(3), 1498e1508.
Pedersen, E. R. (2009). The many and the few: Rounding up the SMEs that manage
CSR in the supply chain. Supply Chain Management, 14(2), 109e116.
Porter, M. E. (1996). What is a strategy? Harvard Business Review, 74(6),
61e78.
Tsai, W. (2001). Knowledge transfer in intraorganizational networks: Effects
of network position and absorptive capacity on business unit innovation
and performance. Academy of Management Journal, 44, 996e1004.
Welford, R., & Frost, S. (2006). Corporate social responsibility in Asian
supply chains. Corporate Social Responsibility and Environmental Man-
agement, 13(3), 166e176.